Alfonso Fausto1, Domenico Rubello2, Alice Carboni3, Paola Mastellari4, Sotirios Chondrogiannis4, Luca Volterrani5. 1. Department of Diagnostic Imaging, Azienda Ospedaliera Universitaria Senese, Policlinico Universitario 'Santa Maria alle Scotte', Viale Bracci 16, 53100 Siena, Italy. 2. Department of Radiology, Nuclear Medicine, Medical Physics, Head Service of Nuclear Medicine, PET/CT Centre, 'Santa Maria della Misericordia Hospital', Via Tre Martiri 8, 45100 Rovigo, Italy. Electronic address: domenico.rubello@libero.it. 3. Department of Medical, Surgical and Neuro Sciences, Diagnostic Imaging, University of Siena, Strada delle Scotte 2, 53100 Siena, Italy. 4. Department of Radiology, Nuclear Medicine, Medical Physics, Head Service of Nuclear Medicine, PET/CT Centre, 'Santa Maria della Misericordia Hospital', Via Tre Martiri 8, 45100 Rovigo, Italy. 5. Department of Diagnostic Imaging, Azienda Ospedaliera Universitaria Senese, Policlinico Universitario 'Santa Maria alle Scotte', Viale Bracci 16, 53100 Siena, Italy; Department of Radiology, Nuclear Medicine, Medical Physics, Head Service of Nuclear Medicine, PET/CT Centre, 'Santa Maria della Misericordia Hospital', Via Tre Martiri 8, 45100 Rovigo, Italy; Department of Medical, Surgical and Neuro Sciences, Diagnostic Imaging, University of Siena, Strada delle Scotte 2, 53100 Siena, Italy.
Abstract
OBJECTIVES: To evaluate ultrasound elastography (USE) using strain ratio (SR), a relative quantification approach for breast lesions characterization. METHODS: One hundred forty-seven consecutive patients with a total of 156 breast lesions underwent USE. Technical accuracy was assessed automatically. For SR evaluation a rounded ROI was depicted inside fat (F), glandular tissue (G) and inside the lesion (L), preferably at the same depth. R1, mean value of the G and F ratio, stands for in background tissue composition elasticity. R2; mean value of L/F stands for in lesion elasticity, both evaluated in arbitrary unit (au). Two-years follow-up and pathology results were standard of reference. Mann-Whitney test, ROC analysis and Chi-square with Yates correction were used. RESULTS: With the exception of 27 cysts, 17 malignant and 112 benign lesions were found. R1 values were 1.6±0.7au and 1.2±0.9au (p=0.015 NS); R2 values were 6.1±2.5au and 1.9±1.3au (p<0.001) for malignant and benign lesions, respectively. A threshold of 3.3au showed a sensitivity and specificity of 88% and 87%, respectively with an AUC of 93%. Fifteen false positive and two false negative were detected. CONCLUSION: Relative quantification of ultrasound elastography allows to find high levels of diagnostic accuracy in characterizing breast tumors above all in downgrading BI-RADS 3 and 4 lesions.
OBJECTIVES: To evaluate ultrasound elastography (USE) using strain ratio (SR), a relative quantification approach for breast lesions characterization. METHODS: One hundred forty-seven consecutive patients with a total of 156 breast lesions underwent USE. Technical accuracy was assessed automatically. For SR evaluation a rounded ROI was depicted inside fat (F), glandular tissue (G) and inside the lesion (L), preferably at the same depth. R1, mean value of the G and F ratio, stands for in background tissue composition elasticity. R2; mean value of L/F stands for in lesion elasticity, both evaluated in arbitrary unit (au). Two-years follow-up and pathology results were standard of reference. Mann-Whitney test, ROC analysis and Chi-square with Yates correction were used. RESULTS: With the exception of 27 cysts, 17 malignant and 112 benign lesions were found. R1 values were 1.6±0.7au and 1.2±0.9au (p=0.015 NS); R2 values were 6.1±2.5au and 1.9±1.3au (p<0.001) for malignant and benign lesions, respectively. A threshold of 3.3au showed a sensitivity and specificity of 88% and 87%, respectively with an AUC of 93%. Fifteen false positive and two false negative were detected. CONCLUSION: Relative quantification of ultrasound elastography allows to find high levels of diagnostic accuracy in characterizing breast tumors above all in downgrading BI-RADS 3 and 4 lesions.
Authors: Hyo Jin Kim; Sun Mi Kim; Bohyoung Kim; Bo La Yun; Mijung Jang; Yousun Ko; Soo Hyun Lee; Heeyeong Jeong; Jung Min Chang; Nariya Cho Journal: Sci Rep Date: 2018-04-18 Impact factor: 4.379